luc blassel

Postdoc▪️Machine learning▪️Bioinformatics - he/him
Come check me out on 🦋 @lblassel.bsky.social

[luc blassel @lucblassel]

We're excited to be sharing our work on end-to-end likelihood-free phylogenetic inference using deep learning with Phyloformer 2.

Check out the preprint if that interests you:
https://t.co/qZIQtszBqi

Or our explanatory thread on bluesky:
https://t.co/UKIseb9dpA

[Rayan Chikhi @RayanChikhi]

#genomeinformatics2024 starting, Camille Marchet (@CamilleMrcht) gives the first keynote on de Bruin Graphs and applications. Conf program https://t.co/2E8Z0jZVtv

[Rayan Chikhi @RayanChikhi]

Next, Roland Faure (@rfaure2) on a new way to sketch sequences while mostly preserving SNPs, using mapping-friendly reductions from @pashadag, @lucblassel and I (Sina's thread: https://t.co/I5E4o98tET)

[Rayan Chikhi @RayanChikhi]

Today we’re excited to freely share an early-version of, perhaps, the world’s most expansive genetics dataset: Logan.

#bioinformatics #petabase #genetics #genomics #openscience

https://t.co/CQSBmvn7se

[luc blassel @lucblassel]

We're finally sharing our work on Phyloformer: a transformer-based approach for likelihood-free phylogenetic reconstruction. Here's a nice thread by @ljacob that recaps it all. If you're curious for more, feel free to read the preprint linked in the thread!

[thequeenfatima @fatimainparis]

avoid @Hotwire like the plague! I booked a hotel reservation with them over the phone and they wrote down the wrong name! I couldn't check in & it was midnight. I spent 4 hours stranded in a new city going back and forth to hotels & customer service. Still waiting 4 a full refund

[luc blassel @lucblassel]

@iskander I am also definitely interested if you still have some :) I’m trying to leave this place

[luc blassel @lucblassel]

Learning about reproducibility in bioinformatics with @fredebibs at #ISMBECCB2023

[Elodie Laine @LaineElodie]

Our next #LEGO research day will take place in Lille, France, on November 23! The event is part of the @CNRS GDR BIM days. You can contribute an abstract until October 20 at: https://t.co/eGjpVnFYxA. #bioinformatics @ljacob @florajay_ @RaphaelMourad @CRouxEvo

[Frederic Lemoine @fredebibs]

ISMB/ECCB 2023 proceedings are out! https://t.co/PxdKYywTIL Editorial by @YannPonty and @sroyyors #ISMBECCB2023 #ComputationalBiology #Bioinformatics #Conference

[Elodie Laine @LaineElodie]

📢Job alert📢
You are looking for a post-doc in deep learning for genomics? We are recruiting!
https://t.co/QNuhkEp47m
#centerofparis @Sorbonne_Univ_ @LCQB_UMR7238 @IBPS_Paris

[Francesco Andreace @frnandreac]

The construction of large cohorts human pangenome graphs is a challenging task. We selected 5 different state-of-the art tools that generate variation graphs or De Bruijn graphs and used them to construct a human pangenome of 104 high-quality human haplotype assemblies. (1/4)

[luc blassel @lucblassel]

@NeuroKaiii @AcademicChatter A little late to the party but here's mine if you want to take a look: https://t.co/RJtw8BFvLy

[Yoann Dufresne @yoann_dufresne]

With plenty of colleagues, we are happy to introduce bioconvert, a software to convert bioinformatics files from one format to another. bioconvert currently includes 50 formats and 100 converters.

[Frederic Lemoine @fredebibs]

The Institut Pasteur Bioinfo/Biostats Hub @HubPasteur is recruiting for research engineer positions, including one focused on viral genomics data analysis ! https://t.co/F5pCChaVHM

[luc blassel @lucblassel]

MSRs are a cheap way to get more accurate mappings for @nanopore data. So read the whole story at: https://t.co/gUyG6YNOp7 and let us know what you think!

9/9

[luc blassel @lucblassel]

Can we do better than homopolymer compression when mapping long @nanopore reads? If you missed #RECOMBSeq2022, find out by reading our @iScience_CP paper: https://t.co/gUyG6YNOp7 with @RayanChikhi and @pashadag.

1/9

[luc blassel @lucblassel]

On repeated regions of the #T2T CHM13 reference, MSRs still improve mapping performance compared to HPC or raw. On centromere-like references though, it is better to not preprocess reads at all, even with HPC.

8/9